Fas (CD95) Expression is Up-Regulated on PapillaryThyroid Carcinoma*
PATRICIA L. ARSCOTT, THEOPHIL STOKES, ANDRZEJ MYC,THOMAS J. GIORDANO, NORMAN W. THOMPSON, AND JAMES R. BAKER, JR.
Departments of Internal Medicine (P.L.A., T.S., A.M., J.R.B.), Pathology (T.J.G.), and Surgery(N.W.T), University of Michigan, Ann Arbor, Michigan 48109
ABSTRACTThyrocyte apoptosis signaled through the Fas receptor has been
proposed as a mechanism for the cytotoxicity observed in thyroiditis,but the role the Fas pathway plays in thyroid cancer is not known. Weexamined Fas expression in thyroid tissue derived from patients withpapillary carcinoma and follicular cancer. More intense immunohis-tological staining for the Fas protein was observed on papillary cancercells as compared with adjacent normal follicles. To further charac-terize the expression of Fas in papillary cancer, paired normal andcancerous thyroid tissues were obtained at thyroidectomy from sev-eral donors, digested, and placed into cell culture. Messenger RNAwas analyzed by ribonuclease protection assays, and protein was
identified by flow cytometry. Fas expression was detected at levels upto 3-fold higher in cancerous thyrocytes compared with paired normalcells. To determine whether the expressed Fas antigen was func-tional, thyrocytes were treated with a monoclonal IgM anti-Fas an-tibody (clone CH11; Upstate Biotechnology, Inc., Lake Placid, NY) inthe presence of interferon-g and cycloheximide. Whereas both normaland cancerous thyrocytes were induced to die after this treatment, thecancerous thyrocytes were more sensitive to anti-Fas antibody. Thiswork demonstrates that the Fas antigen is expressed and functionalon papillary thyroid cancer cells and this may have potential thera-peutic significance. (J Clin Endocrinol Metab 84: 4246–4252, 1999)
AMONG thyroid tumors, the presence of immune cellsis most often observed in association with papillary
thyroid cancer (PTC) (1). Antibodies to thyroid antigens havebeen documented in the serum of patients with PTC (2), andthyroid-specific, cell-mediated immune responses have alsobeen demonstrated (3). Interestingly, the presence of an an-tithyroid immune response in PTC has been reported to beassociated with a better prognosis (4). PTC also has beenassociated with autoimmune thyroid diseases, and patientswith concurrent thyroiditis have better survival rates (1, 5, 6).However, the basis for the association of an immune re-sponse to PTC with improved outcome is unknown.
Lymphocytic infiltrates in thyroid cancer contain cytotoxicT lymphocytes (CTL) (7, 8), and a major mechanism of targetcell lysis by CTL is through Fas-mediated apoptosis (9). Inthis process, Fas ligand (FasL) expressed on CTL binds to Fason tumor cells, initiating an apoptotic signal (10). Fas hasbeen detected on some tumors and on several tumor celllines, including pancreas (11), lung (12), and gastric mucosa(13). In contrast, it has been suggested that tumors that do notexpress Fas antigen escape immune surveillance (14).
Whether Fas-signaled apoptosis is involved in immunecontrol of PTC is unknown. Previously published work hasshown that Fas is expressed on normal thyroid follicular cellsand may play a role in the cellular destruction seen in thy-roiditis (15). In cases of differentiated thyroid cancer, how-ever, the presence and function of Fas is unknown.
In this study, we have examined the expression of Fas onthyroid cancer cells. We found that Fas was up-regulated inmany papillary cancers compared with normal thyroid fol-licles. We then evaluated the potential for activation of theFas death pathway in vitro and found that PTC thyrocytesalso display increased sensitivity to the induction of apo-ptosis through this mechanism.
Materials and MethodsImmunostaining of thyroid tissue sections
Formalin-fixed, paraffin-embedded sections from different types ofthyroid cancer were obtained after pathological examination and im-munostained for the presence of Fas antigen. Each slide was deparaf-finized with three rinses of xylene, followed by rehydration with eth-anol, and finally rinsed with phosphate-buffered saline (PBS). Tounmask antigenic determinants, slides were microwave-pretreated for15 min in 0.01 m citrate buffer, then washed with PBS and blocked with5% normal goat serum in PBS for 20 min. The slides were then incubatedfor 4 h with either rabbit anti-Fas polyclonal antibody (N-18; Santa CruzBiotechnology, Inc., Santa Cruz, CA) or rabbit IgG control antibody(Jackson ImmunoResearch Laboratories, Inc., West Grove, PA) at 1mg/mL in 1.5% normal goat serum. After washing with PBS, the slideswere incubated with biotinylated goat antirabbit IgG and detected usingan avidin-biotin complex detection kit with glucose oxidase substrate(Vectastain ABC-GO kit; Vector Laboratories, Inc. Burlingame, CA).Stained slides were briefly counterstained with eosin and mounted withpermount (Sigma Chemical Co., St. Louis, MO).
Thyroid cell culture
To examine differences in Fas expression and function, thyroid tissuewas obtained at the time of thyroidectomy from seven patients with PTC.Normal and cancerous tissue were separated at surgery and were treatedin parallel. The tissue was digested overnight with 40 mg collagenase, 4 mghyaluronidase, and 4 mg DNase I (all from Sigma Chemical Co.) in 40 mLRPMI 1640 (Life Technologies, Inc., Grand Island, NY). Red blood cells werelysed with ammonium chloride lysis buffer [0.15 m NH4Cl, 10 mm KPO4,and 1 mm EDTA (pH 7.3)] and cells were cultured in Cellgro Complete
Received June 16, 1999. Accepted August 13, 1999.Address correspondence and requests for reprints to: James R. Baker,
Jr., M.D., University of Michigan Medical Center, 1150 West MedicalCenter Drive, Room 9220 MSRB III, Ann Arbor, Michigan 48109-0648.E-mail: [email protected].
Media (Mediatech, Herndon, VA) with 20% NuSerum IV (CollaborativeBiomedical Products, Becton Dickinson and Co. Labware, Bedford, MA) at37 C in 5% CO2. After 24 h, nonadherent cells were removed by washingwith media. Adherent thyrocytes were supplemented with 10 mIU/mLbovine TSH (Sigma Chemical Co.) every 2–3 days. Cells were stained forthyroglobulin to ensure that they were thyroid in origin (16).
Ribonuclease protection assay
RNA was isolated from thyrocytes using TriZol (Molecular ResearchCenter, Inc., Cincinnati, OH). RiboQuant™ Multi-Probe RNase Protec-tion Assay System (PharMingen, San Diego, CA) was used for thedetection and quantitation of specific messenger RNA (mRNA) species.[32P]-labeled antisense RNA probes were prepared using the HumanApoptosis hAPO-3 Template Set (PharMingen), which included humanFas, FasL, Fas-associated phosphatase-1 (FAP-1), and glyceraldehyde3-phosphate dehydrogenase (GAPDH). The probes were hybridizedwith 10 mg total thyrocyte RNA, 2 mg HeLa RNA (assay-positive con-trol), and 2 mg yeast transfer RNA (background control). After hybrid-ization, the samples were subjected to RNase treatment, followed bypurification of RNase-protected probes. The protected probes were re-solved on a 5% denaturing polyacrylamide gel. The quantity of specifictranscript present was analyzed by autoradiography and densitometricanalysis of scanned films using Quantity One software (Bio-Rad Lab-oratories, Inc., Hercules, CA). Relative amounts of specific message werecorrected for RNA loading by comparing with the GAPDH band in-tensity for each sample.
Flow cytometry analysis
Normal and cancer-derived thyrocytes were lifted from 10-cm culturedishes using trypsin/ethylenediaminetetraacetate (Life Technologies,
Inc.). After washing with PBS, 1–2 3 106 cells/mL were fixed with 1%formaldehyde in PBS for 30 min on ice. Cells were pelleted and per-meabilized with 0.1% Triton X-100 in PBS with 0.1% BSA (PBA). Afterwashing with PBS, cells were dual-stained with a mouse anti-Fas mono-clonal antibody (mAb) (clone UB2; MBL International, Watertown, MA)and a rabbit antithyroglobulin polyclonal antibody (DAKO Corp.,Carpinteria, CA). Controls included cells stained using isotype controlantibodies, as well as single-stained cells. After a 1-h incubation, cellswere washed with PBA, then incubated with fluorescein isothiocyanate-conjugated antimouse IgG and phycoerythrin-conjugated antirabbit IgG(Jackson ImmunoResearch Laboratories, Inc.) for 1 h. Finally, cells werewashed with PBA and fluorescence was detected with a FACScan flowcytometer (Becton Dickinson and Co., San Jose, CA) and analyzed withCellQuest software (Becton Dickinson and Co., San Jose, CA).
Induction of apoptosis
Cultured thyrocytes were pretreated with and without 500 IU/mLhuman interferon (IFN)-g for 48 h. To initiate apoptosis, thyrocytes werethen treated for 24 h with 0.2–0.8 mg/mL mouse IgM anti-Fas mAb, cloneCH11 (Upstate Biotechnology, Inc., Lake Placid, NY) or purified mouseIgM (Sigma Chemical Co.) as a control. Some cells were also treated withconcentrations of cycloheximide (CHX) that do not yield complete ap-optosis, from 2.5–5 mg/mL, concomitantly with the mAb. Apoptosis wasevaluated by observation of morphological changes, and cell death wasmeasured using a 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazoliumbromide (MTT) assay.
MTT quantitation of cell viability
Cultured thyrocytes were grown in 96-well plates at cell densities of10,000 - 20,000 cells/well in 0.1 mL and were treated to induce apoptosis,
FIG. 1. In vivo Fas expression in thyroid cancer was evaluated by immunohistochemical staining of thyroid sections using a rabbit polyclonalantibody to the amino terminus of Fas. Three different PTC cases (A-C) and a follicular variant of PTC (D) show normal thyroid follicles adjacentto a papillary tumor. Normal follicular cells reveal low level Fas expression, whereas papillary cancer cells demonstrated a strong increase instaining for Fas (A, arrows).
FAS EXPRESSION IN PAPILLARY CARCINOMA 4247
as described. Metabolically active cells were then detected by addingMTT to a final concentration of 0.5 mg/mL for 2–4 h at 37 C. Isopropanolwith 0.04N HCl was added, wells were mixed, and the plates were readat 595 nm with a 630-nm reference (17). Mean and se for triplicate wellsare reported.
Statistics
Differences between anti-Fas mAb or IgM control-treated thyrocyteswere compared using a paired two-tailed t test. Differences were con-sidered significant when P , 0.05.
Results
Fas expression in thyroid sections
In vivo Fas expression in thyroid cancer was evaluated byimmunohistochemical staining of thyroid sections using arabbit polyclonal antibody to the amino terminus of Fas(N-18). Thyroid cancer specimens from 18 patients were ex-amined and included seven PTC specimens, four follicularvariants of PTC, and seven follicular cancer (FC) specimens.
FIG. 2. Fas protein expression onpaired normal and cancer thyrocyteswas compared by flow cytometric anal-ysis using a mAb against Fas. A, Fasfluorescence for both normal (light his-togram) and cancer (dark histogram)thyrocytes is plotted along the x-axis, onlog scale. B, The mean channel fluores-cence for each cell population after sub-traction of background fluorescence iscompared for paired normal (M) andcancer (f) thyrocytes from two differentPTC patients.
4248 ARSCOTT ET AL. JCE & M • 1999Vol 84 • No 11
The interpretation of staining on normal and cancerous thy-roid cells was evaluated independently by two individuals.Some PTC regions displayed intense Fas staining over broadareas of tumor, whereas others showed either patchy stain-ing, or less intense staining. Three specimens did not containa region of normal follicles for comparison. The thyroid spec-imens shown in Fig. 1 demonstrate differential staining ofnormal thyroid follicles adjacent to a papillary tumor. Al-though normal follicular cells revealed low-level Fas expres-sion, the tumor cells demonstrated a higher density of Fasantigen (Fig. 1A, arrows). A substantial increase in Fas ex-pression in tumor cells compared with adjacent normal thy-roid follicles was demonstrated in three PTC specimens andthree follicular variants of PTC. Five of seven FC specimensdemonstrated low levels of Fas expression, and in only onespecimen was a slight increase observed in the tumor cellscompared with normal regions.
Fas expression in normal and cancer thyrocytes
We examined Fas expression on paired normal and cancer-derived thyrocytes from the same patient to see if the dif-ferences observed in vivo were maintained in culture. Fasprotein expression on thyrocytes isolated from two patientswith PTC was compared by flow cytometric analysis usinga mAb against Fas (clone UB2). Fas was expressed on morethan 90% of normal thyrocytes as compared with isotypecontrol stained cells. However, a shift in the amount ofFas-specific fluorescence was evident in the cancer cell pop-ulation (Fig. 2A). For both patients, the mean channel fluo-rescence of anti-Fas-stained cells was increased in cancer-derived thyrocytes compared with normal thyrocytes (Fig.2B). Fas expression was confirmed by Western blot for thy-rocytes from two other patients, however, no difference inexpression between normal and cancer thyrocytes could bemeasured using this technique (data not shown).
Evidence for transcriptional differences in Fas expres-sion was provided by ribonuclease protection assay todetect Fas mRNA in thyrocytes from patients with PTC.We compared the intensity of the band representing Fasmessage, normalized to the intensity of the band forGAPDH for each sample (Fig. 3). Fas mRNA was present
in both normal and cancerous thyrocytes from all fourpatients. In two of the four paired samples (Fig. 3, sets 2and 4), the cancer-derived thyrocytes demonstrated levelsof Fas mRNA that were substantially higher than in nor-mal thyrocytes from the same patient. In the two othercases, differences were negligible.
Activation of Fas-mediated cell death in normal andcancer-derived thyrocytes
To determine whether increased Fas expression on thyroidcancer cells could affect the propensity of thyrocytes to un-dergo Fas-mediated apoptosis, paired thyrocytes weretreated with a control mouse IgM antibody or an anti-FasmAb (clone CH11), which cross-links the Fas receptor, ini-tiating a death signal. When treated with anti-Fas mAb alone,no cell death was observed in either normal or cancerousthyrocytes as measured by an MTT cell viability assay (Fig.4A, left). Because earlier studies demonstrated that a labileinhibitor of the Fas pathway rendered thyrocytes resistant toapoptosis by the anti-Fas mAb alone (16), CHX was thenincluded with mAb treatment. Cancer thyrocytes treatedwith anti-Fas mAb in the presence of CHX demonstrated a47% reduction in cell viability compared with IgM controlantibody-treated cells (Fig. 4A, right) and were more sensi-tive to death induced in this manner than similarly treatednormal thyrocytes (Fig. 4A, left).
Pretreatment of thyrocytes with IFN-g for 48 h can enhancenormal thyrocyte susceptibility to Fas-mediated apoptosis(16). This was also observed in the cancer thyrocytes wherethe death induced by anti-Fas mAb and CHX was signifi-cantly increased after exposure to IFN-g (Fig. 4B). An evengreater increase in cell death was observed in the corre-sponding normal thyrocytes, to the point where the responsein both cell populations was similar.
Normal and cancer thyrocytes from five PTC patients werecompared for sensitivity to Fas-mediated cell death, as mea-sured by MTT assay. Fas-mediated death in normal thyro-cytes ranged from 6–49%, whereas cancer thyrocytes dis-played 20–78% death. In four of the five cases, the cancerthyrocytes were more sensitive to death induced by anti-FasmAb in the presence of CHX, demonstrating 1.5–7 timesmore death than observed in normal thyrocytes.
FAP-1 mRNA in normal and cancer thyrocytes in vitro
The requirement of CHX for both normal and cancerousthyrocytes to respond to Fas-mediated apoptosis suggeststhat inhibitors of this pathway may also play a role in de-termining sensitivities to death induced by treatment withanti-Fas mAb. We examined the expression of one such in-hibitor, FAP-1 (18), which has been reported to be present inthe thyroid (19). The level of mRNA for FAP-1 was analyzedby ribonuclease protection assay in paired normal and cancerthyrocytes from four patients (Fig. 5). Interestingly, levels ofFAP-1 did not differ substantially between the normal andcancer thyrocytes, even in the two sets where Fas mRNA wasincreased in cancer thyrocytes (Fig. 3, sets 2 and 4).
FIG. 3. Levels of Fas mRNA were compared in paired thyrocytes fromfour patients using a ribonuclease protection assay with radiolabeledprobes specific for Fas and the housekeeping gene GAPDH. The in-tensity of the band representing Fas message was normalized to theintensity of the bands for GAPDH and is expressed as a ratio ofrelative band intensity (indicated below each sample).
FAS EXPRESSION IN PAPILLARY CARCINOMA 4249
Discussion
Constitutive expression of Fas antigen has been demon-strated in normal thyroid, as well as in Hashimoto’s thy-roiditis and other thyroid disorders (15, 16, 20–22). Our re-sults demonstrate that Fas is also expressed in PTC. Thepresent study, however, extends these findings by demon-strating that, in some cases of PTC, Fas expression is in-creased on tumor cells compared with surrounding normaltissue. Specimens from both PTC and FC were examined, butonly PTC demonstrated higher levels of Fas expression com-pared with surrounding normal tissue. Both PTC and FC aredifferentiated tumors, although considered distinct tumortypes. PTC is more common, but has a better 10-yr prognosisthan FC (1). Both PTC and FC are linked to radiation expo-sure, although ras is believed to play a primary role in tumortransformation in FC (23) and ret/PTC appears almost ex-clusively in PTC (24, 25).
In thyroid cells associated with thyroiditis, Fas is consti-
FIG. 5. Levels of mRNA for FAP-1 were compared in paired thyro-cytes from the same four patients shown in Fig. 3, using a ribonu-clease protection assay with radiolabeled probes specific for FAP-1and the housekeeping gene GAPDH. The intensity of the band rep-resenting FAP-1 message was normalized to the intensity of the bandsfor GAPDH and is expressed as a ratio of relative band intensity(indicated below each sample).
FIG. 4. Activation of Fas-mediated celldeath in normal and cancer thyrocytes.A, Normal and cancer thyrocytes from arepresentative patient were induced todie with anti-Fas mAb (f) or treatedwith control antibody (u) in the pres-ence or absence of a suboptimal concen-tration of CHX. Cell viability was mea-sured using MTT. Each value is theaverage and SE from three wells (*P ,0.005 vs. IgM control). B, Thyrocytestreated with the same combinations ofantibodies and CHX after pretreatmentwith IFN-g.
4250 ARSCOTT ET AL. JCE & M • 1999Vol 84 • No 11
tutively expressed and apoptosis is also evident, especiallyin regions proximal to lymphocytic centers (15, 26). The Fasexpressed in Hashimoto’s thyroiditis is, therefore, thought tocontribute to increased apoptosis in the thyroid (15, 21). PTCis often accompanied by lymphocytic infiltrates (4), and ev-idence of increased levels of apoptosis in PTC has also beenreported (15, 27). Thus, the Fas expressed on PTC may pro-vide a possible mechanism for the initiation of apoptosis inneoplastic thyroid cells.
We then considered whether the increased expression ofFas in PTC is functional and, therefore, a potential target forimmune control. By examining thyrocytes isolated from nor-mal and cancerous tissue from the same patient, we were ableto directly compare the expression of Fas, and the ability ofthese cells to respond to Fas-mediated cell death. Expressionof Fas mRNA and protein were documented in both normaland cancer-derived thyrocytes. Whereas Fas mRNA levelsdid not differ between normal and cancerous tissue for allpatients, the increased levels found in the cancer thyrocytesfrom two patients suggest possible transcriptional control inthese patients. Importantly, it is the expression of Fas proteinthat more directly relates to the susceptibility of PTC thy-rocytes to Fas-mediated apoptosis. As quantified by flowcytometry, Fas protein expression on cancer-derived thyro-cytes was increased nearly 3-fold compared with normalthyrocytes. Correspondingly, susceptibility to Fas-mediateddeath was also significantly enhanced in these PTC cells.
We have previously shown that thyrocytes were resistantto death initiated by anti-Fas mAb alone, but susceptible inthe presence of CHX, which reduced levels of a labile in-hibitor (16). In our present study, the cancer-derived thyro-cytes, as well as the normal thyrocytes, also demonstrated thepresence of a labile inhibitor of Fas-mediated apoptosis andrequired the presence of CHX to initiate death. In the case ofPTC, once inhibitor levels were decreased by CHX it is likelythat the up-regulation of Fas on cancer thyrocytes was thena significant factor contributing to increased sensitivity toFas-mediated cell death. After IFN-g pretreatment, anti-FasmAb in the presence of CHX induced significant death, inboth normal and cancer thyrocytes, demonstrating that bothcell populations are capable of responding with maximaldeath.
The presence of inhibitors of Fas-mediated apoptosis maycontribute to the overall sensitivity of thyrocytes to deathsignaled through this pathway. It is probable that activationof the Fas death pathway in thyrocytes depends on the rel-ative concentrations of Fas and Fas inhibitor present. Inhib-itors that specifically block Fas signaling include the FLICE(FADD-like ICE)-inhibitory protein (28) and FAP-1 (18),which binds to the negative regulatory domain of Fas. Wehave detected FLICE-inhibitory protein in thyrocytes, how-ever, its expression was not regulated by CHX treatment(data not shown) and, therefore, is unlikely to be directlyinvolved in thyrocyte resistance to death induced by Fassignaling. On the other hand, FAP-1 is present in normalthyrocytes, and protein expression decreases after CHX treat-ment (19), so it may contribute to resistance to Fas-mediatedapoptosis observed in thyrocytes. We have provided evi-dence that FAP-1 is also expressed by thyrocytes in PTC.Although, Fas mRNA is present at higher levels in some
cancer thyrocytes, little difference was seen in the levels ofinhibitor mRNA. Thus, the differential sensitivity of cancerand normal thyrocytes to Fas-mediated cell death is lesslikely the result of depressed inhibitor concentrations thanthe level of Fas expressed on these cells.
The Fas/FasL system has been proposed as a mechanismfor tumor cell defense (29), although there is controversyregarding the importance of FasL in this role (30). Interest-ingly, using ribonuclease protection assay to detect specificFasL mRNA, we found no expression of FasL in either nor-mal or cancer thyrocytes (31) (data not shown), whereasmRNA for Fas was present in all cases. Unfortunately, con-troversy over the specificity of several antibodies to FasLmake the study of FasL protein expression difficult (31–33).Conversely, the Fas expressed on PTC may allow for theimmune control of tumor growth and provide a possibletarget for treatment. Death of tumor cells has been demon-strated after inducing sensitivity to Fas-mediated apoptosiswith anticancer drugs, such as doxorubicin (34). In anotherstudy, treatment with IFN-a induced expression of func-tional CD95 in basal cell carcinomas, eventually leading totumor regression (35). Thus, similar treatments or factors thatenhance Fas-mediated apoptosis could effect PTC tumorlysis.
In summary, our results indicate that Fas is up-regulatedin some PTC cases, and these cells demonstrate greater sen-sitivity to apoptosis through this death pathway. This, alongwith the improved outcome in the presence of immune in-filtration, suggests that the Fas death pathway could be in-volved in the immune control of PTC. Furthermore, ourresults suggest that the increased Fas expressed on some PTCmay provide a possible target for treatment.
Acknowledgments
We thank Dr. James D. Bretz for many helpful discussions.
References
1. Baker JR Jr, Fosso CK. 1993 Immunological aspects of cancers arising fromthyroid follicular cells. Endocr Rev. 14:729–746.
2. Pacini F, Mariotti S, Formica N, et al. 1988 Thyroid autoantibodies in thyroidcancer: incidence and relationship with tumor outcome. Acta Endocrinol.119:373–380.
3. Juhasz F, Boros P, Szegedi G, et al. 1989 Immunogenetic and immunologicstudies of differentiated thyroid cancer. Cancer. 63:1318–1326.
4. Matsubayashi S, Kawai K, Matsumoto Y, et al. 1995 The correlation betweenpapillary thyroid carcinoma and lymphocytic infiltration in the thyroid gland.J Clin Endocrinol Metab. 80:3421–3424.
5. Segal K, Ben-Bassat M, Avraham A, Har-El G, Sidi J. 1985 Hashimoto’sthyroiditis and carcinoma of the thyroid gland. International Surgery.70:205–209.
6. McConahey WM, Hay ID, Woolner LB, van Heerden JA, Taylor WF. 1986Papillary thyroid cancer treated at the Mayo Clinic, 1946 through 1970: initialmanifestations, pathologic findings, therapy, and outcome. Mayo Clin Proc.61:978–996.
7. Bagnasco M, Venuti D, Paolieri F, Torre G, Ferrini S, Canonica GW. 1989Phenotypic and functional analysis at the clonal level of infiltrating T lym-phocytes in papillary carcinoma of the thyroid: prevalence of cytolytic T cellswith natural killer-like or lymphokine-activated killer activity. J Clin Endo-crinol Metab. 69:832–836.
8. Iwatani Y, Hidaka Y, Matsuzuka F, Kuma K, Amino N. 1993 Intrathyroidallymphocyte subsets, including unusual CD41 CD81 cells and CD3loTCR ab lo/2CD42CD82 cells, in autoimmune thyroid disease. Clin Exp Immunol.93:430–436.
9. Lowin B, Hahne M, Mattmann C, Tschopp J. 1994 Cytolytic T-cell cytotoxicityis mediated through perforin and Fas lytic pathways. Nature. 370:650–652.
10. Nagata S, Golstein P. 1995 The Fas death factor. Science. 267:1449–1456.11. Ungefroren H, Voss M, Jansen M, et al. 1998 Human pancreatic adenocar-
FAS EXPRESSION IN PAPILLARY CARCINOMA 4251
cinomas express Fas and Fas ligand yet are resistant to Fas-mediated apoptosis.Cancer Res. 58:1741–1749.
12. Hellquist HB, Olejnicka B, Jadner M, Andersson T, Sederholm C. 1997 Fasreceptor is expressed in human lung squamous cell carcinomas, whereas bcl-2and apoptosis are not pronounced: a preliminary report. Br J Cancer.76:175–179.
13. Hayashi H, Tatebe S, Osaki M, Goto A, Suzuki Y, Ito H. 1997 Expression ofFas antigen and its mediation of apoptosis in human gastric cancer cell lines.Jpn J Cancer Res. 88:49–55.
14. Gratas C, Tohma Y, Barnas C, Taniere P, Hainaut P, Ohgaki H. 1998 Up-regulation of Fas (APO-1/CD95) ligand and down-regulation of Fas expres-sion in human esophageal cancer. Cancer Res. 58:2057–2062.
15. Tanimoto C, Hirakawa S, Kawasaki H, Hayakawa N, Ota Z. 1995 Apoptosisin thyroid diseases: a histochemical study. Endocr J. 42:193–201.
16. Arscott PL, Knapp J, Rymaszewski M, et al. 1997 Fas (APO-1, CD95)-mediatedapoptosis in thyroid cells is regulated by a labile protein inhibitor. Endocri-nology. 138:5019–5027.
17. Mosmann T. 1983 Rapid colorimetric assay for cellular growth and survival:application to proliferation and cytotoxicity assays. J Immunol Methods.65:55–63.
18. Sato T, Irie S, Kitada S, Reed JC. 1995 FAP-1: a protein tyrosine phosphatasethat associates with Fas. Science. 268:411–415.
19. Myc A, Arscott PL, Baker JR Jr. 1998 FAP-1 is present in human thyroidfollicular cells and functions as an inhibitor of Fas-mediated signaling ofprogrammed cell death. Proc 71st Meeting of The American Thyroid Assoc.,Portland, OR, 1998, p 148.
20. Kawakami A, Eguchi K, Matsuoka N, et al. 1996 Thyroid-stimulating hor-mone inhibits Fas antigen-mediated apoptosis of human thyrocytes in vitro.Endocrinology. 137:3163–3169.
21. Giordano C, Stassi G, De Maria R, et al. 1997 Potential involvement of Fasand its ligand in the pathogenesis of Hashimoto’s thyroiditis. Science.275:960–963.
22. Hammond LJ, Lowdell MW, Cerrano PG, Goode AW, Bottazzo GF, MirakianR. 1997 Analysis of apoptosis in relation to tissue destruction associated withHashimoto’s autoimmune thyroiditis. J Pathol. 182:138–144.
23. Farid NR, Shi Y, Zou M. 1994 Molecular basis of thyroid cancer. Endocr Rev.15:202–232.
24. Santoro M, Carlomagno F, Hay ID, et al. 1992 Ret oncogene activation inhuman thyroid neoplasms is restricted to the papillary cancer subtype. J ClinInvest. 89:1517–1522.
25. Farid NR, Zou M, Shi Y. 1995 Genetics of follicular thyroid cancer. EndocrinolMetab Clin North Am. 24:865–883.
26. Kotani T, Aratake Y, Hirai K, Fukazawa Y, Sato H, Ohtaki S. 1995 Apoptosisin thyroid tissue from patients with Hashimoto’s thyroiditis. Autoimmunity.20:231–236.
27. Okayasu I, Saegusa M, Fujiwara M, Hara Y, Rose NR. 1995 Enhanced cellularproliferative activity and cell death in chronic thyroiditis and thyroid papillarycarcinoma. J Cancer Res Clin Oncol. 121:746–752.
28. Irmler M, Thome M, Hahne M, et al. 1997 Inhibition of death receptor signalsby cellular FLIP. Nature. 388:190–195.
29. Hahne M, Rimoldi D, Schroter M, et al. 1996 Melanoma cell expression of Fas(Apo-1/CD95) ligand: implications for tumor immune escape. Science.274:1363–1366.
30. Chappell DB, Restifo NP. 1998 T cell-tumor cell: a fatal interaction? CancerImmunol Immunother. 47: 65–71.
31. Stokes TA, Rymaszewski M, Arscott PL, et al. 1998 Constitutive expressionof FasL in thyrocytes. Science. 279:2015a.
32. Fiedler P, Schaetzlein CE, Eibel H. 1998 Constitutive expression of FasL inthyrocytes. Science. 279:2015a.
33. Smith D, Sieg S, Kaplan D. 1998 Technical note: aberrant detection of cellsurface Fas ligand with anti-peptide antibodies. J Immunol. 160:4159–4160.
34. Mizutani Y, Okada Y, Yoshida O, Fukumoto M, Bonavida B. 1997 Doxoru-bicin sensitizes human bladder carcinoma cells to Fas-mediated cytotoxicity.Cancer. 79:1180–1189.
35. Buechner SA, Wernli M, Harr T, Hahn S, Itin P, Erb P. 1997 Regressionof basal cell carcinoma by intralesional interferon-a treatment is mediatedby CD95 (Apo-1/Fas)-CD95 ligand-induced suicide. J Clin Invest. 100:2691–2696.
